This competitive renewal application has an overall goal, the creation of an atlas of Alzheimer's disease. The neuroscience and informatics efforts proposed here will result in a tool set and product that is applicable not only to the basic and clinical science of Alzheimer's disease, but to the general problem of mapping the structure and function of any dynamic process in health or disease in whole populations of subjects. Leveraging the accomplishments achieved during the last period of this project and building upon our high-resolution post mortem anatomic framework, the development of atlas construction methodology and the ability to create 3D visual models of anatomy, we will construct the first multimodality probabilistic atlas of the brain representing a diseased population. Including both histologically processed post mortem tissue as well as high- resolution 3D MR images acquired from subjects in various stages of Alzheimer's disease, we will generate the average geometry and 3D variability of the anatomic structures of these populations. Further, we will describe the anatomy as cytoarchitectural features from histology and gyral sulcal features from MRI. There are 7 specific aims in this project. The first will be the collection of a cohort of post mortem specimens from an Alzheimer's disease population. Second, we will create detailed individual probabilistic maps describing the architectural boundaries in AD and matched controls. Third, we will create an MRI probabilistic atlas based upon data that has been previously acquired or will be acquired with funding from other active projects. Fourth, we will develop and refine appropriate registration deformation correction atlasing strategies to create a comprehensive multimodality atlas of Alzheimer's disease. This will enable the development of data at different spatial resolutions and representing different aspects of brain structure and function. Fifth, individualized data analysis utilizing mathematical strategies to compare individual MRI data with the probabilistic atlas will enable access by the neuroscience community to this multimodality atlas. Sixth, we will develop dynamic 4D mapping tools to express the spatial and temporal profiles of degeneration heretofore unavailable in static single time point representations of anatomy or physiology. Seventh, these will be combined into an interactive visualizable and analytic tool set made available to the neuroscientific community.